Postgraduate supervision at an open distance e-learning institution in South Africa

Effective postgraduate supervision is a concern at universities worldwide, even under optimal conditions where postgraduate students are studying full-time. Universities are being pressured by their governments to increase the throughput of postgraduates where there is a need for supervisory guidance in order to produce quality graduates within a shorter period of time than was previously thought possible. In an Open Distance E-learning (ODeL) context in South Africa, postgraduate supervision presents an even more formidable task as face-to-face communication between supervisor and student is restricted or totally non-existent. Informed by a review of the Community of Practice Theory, the researcher undertook a qualitative study to investigate the challenges of supervisors by means of a purposeful sample of postgraduate supervisors at a major ODeL institution in South Africa. Open-ended questionnaires were the means of collecting the data. The results indicated weaknesses in respect of the following: the selection and allocation of postgraduate students to supervisors without consultation; the requirements for intensive guidance during the process of writing the thesis to meet the needs of underprepared students; and the difficulties inherent in the ODeL model, which depended primarily on written communication, especially for academically weak students. The recommendation is that experienced supervisors should conduct workshops and mentor novice academics on effective ODeL supervision procedures.


Introduction
The supervision of postgraduate study is a challenge at universities worldwide, even under optimal conditions and where the postgraduate students study full-time (Andrew, 2012;Sussex, 2008;Willems, Farley, Ellis, McCormick & Walker, 2011;Wisker, 2008). However, postgraduate supervision is a global policy imperativeit plays a critical role in the global economy, as the accumulation of knowledge is a factor affecting the productive capacity of a country, hence its global competitiveness. Among several factors, the success of postgraduate study supervision is dependent on a sound relationship between supervisors and supervisees, especially in respect of their cognitive abilities to plan and coordinate their study project in an intelligent manner (Bitzer, 2011;Koen, 2007;Lessing, 2011;Yeatman, 1995). In the past, postgraduate studies were accepted to postgraduate studies primarily by invitation from supervisors, who were allocated only a few students to supervise. However, the massification and marketisation of higher education has resulted in increasing numbers of students entering postgraduate programmes with different levels of capabilities (McCormack, 2012). Coupled with this is the mounting pressure from the governments to deliver postgraduate students within a prescribed period of time in order for these newly qualified individuals to contribute economically to the development of the nation (Manathunga, 2012). The Brazilian government, for example, was able to improve its economic development by increasing the quality and quantity of its postgraduate students (Sandoval, 2012). According to Mouton (2011), South Africa's ability to improve its global competitive edge is dependent on the quality of its postgraduate students. Moreover, since the government's funding to universities is dependent on the students' throughput rate, the supervisors have an added burden to find new ways of guiding these students quickly and effectively.
Compounding the above challenges is the fact that postgraduate supervision often involves the geographical distance between the students and the supervisors within an Open Distance e-learning (ODeL) context. Distance education is a kind of education being offered to students who do not attend classes daily and hence are not in touch with their lecturers and supervisors (Wisker, Robinson, Trafford, Creighton & Warner, 2003). Distance education has been in existence for centuries, with developed countries such as the United States of America, Australia and the United Kingdom being in the forefront (University of London (UOL), 2012). The University of London first offered postgraduate programmes for master's and doctoral students through distance education in 1969 (UOL, 2012). Other countries followed soon thereafter. For example, Canada established the University of Athabasca in 1972, and Germany established its Open Distance Education in 1974 (UOL, 2012). These programmes were for mid-career students who preferred to conduct research aimed at improving their work performances (UOL, 2012). Recently, many higher education institutions of distance learning address issues of quality and timely completion through dual mode systems, hence the more recent term, ODeL institutions (Guilar & Lorring, 2008;Mouton, 2011). In this regard, both e-learning and contact sessions are on offer in order to assist postgraduate students to complete their studies on time and to produce high quality theses and dissertations.
In South Africa, as in other higher education institutions around the world, a concerted endeavour exists to increase the number of successful postgraduate students. To this end, the National Department of Higher Education and Training (DHET, 2012) has set goals to produce more master's and doctoral graduates in order to improve the country's economic competitiveness. Presently, South Africa has one of the lowest graduation rates at master's and doctoral levels, a rate of less than 15% a year, compared to most developed countries for example, the USA produces 288 doctoral students a year and Britain produces 395 students per year (Mouton, 2011). Notwithstanding the drive for more graduates at these levels, the South African government discourages universities from admitting academically under-prepared students with limited chances of completing their studies, as this becomes an unwise use of scarce financial resources (Department of Education (DoE), 2005). In the past, the South African government's funding formula for higher education was dependent on the intake of the number of students. However, funding now depends on student throughput (DHET, 2012). In terms of ODeL, the Department of Higher Education expects postgraduate students to complete their master's degree within four years and their doctoral degrees within six years (DHET, Republic of South Africa, 2012).
The University of South Africa (UNISA) is the only comprehensive open and distance elearning (ODeL) institution in South Africa with a mixed-mode approach, although e-learning is most common for postgraduate students. According to UNISA (2008:2), ODeL is a multi-dimensional concept, which aspires at bridging the gap between the students and the institution, the students and the academics and the students and their peers. UNISA (2008) aims to increase its doctoral throughput rates by at least 25%, in the light of national needs. This percentage is higher than the figure set by the DHET, Republic of South Africa (2012). The pressure to increase the throughput rate of postgraduate students at UNISA is not unique to South Africa, but is a phenomenon at most institutions around the world (Swanepoel, 2010). In addition, more postgraduate students from diverse cultural backgrounds are now registered students at UNISA (Letseka & Pitsoe, 2014). Most of these students are from previously disadvantaged groups, such as females, and language minority groups. Since these groups are new to the system and most of them lack a culture of research, supervisors' work has become increasingly difficult (Mouton, 2011). Unfortunately, the opening up of access to higher education has not resulted in a change in institutional culture to accommodate these students and to ensure their academic success (Mouton, 2011). Most academics who are appointed to supervise these students often lack the necessary skills and knowledge to work with students from diverse cultural backgrounds (Malan, Erwee, Van Rensburg & Danaher, 2012). As a result, the supervisors as well as the students bring different expectations to the supervision relationship. These different expectations may create conflict, because supervision revolves around the relationship between a supervisor and supervisee. In particular, not much has taken place to change the supervision styles in an ODeL context, supervision is still considered in the same way as it was for fulltime students (Wisker, Robinson & Shacham, 2007). Supervision has retained its historical patronage culture, as developed in face-to-face traditional institutions, and the supervisors rely on their own experience of supervision to supervise their students (Lessing, 2011;Mouton, 2007).
Against this background, a qualitative research study investigated the challenges in postgraduate supervision faced by academics at a major ODeL institution, namely UNISA. The main research question that this study sought to address, was: What are the challenges faced by supervisors at UNISA who are supervising master's and doctoral students through ODeL?
In the next section, the researcher discusses the theoretical framework which underpins this study.

Theoretical Framework
The theoretical framework that underpins this study is the Community of Practice Theory by Etienne Wenger (1999). While studying apprenticeship as a learning model, Lave and Wenger (1999) first coined the term community of practice. Lave and Wenger (1999) investigated how people from informal sector were motivated by a desire to share learning, in addition to their organisation's requirements. According to Wenger (1999:10), a 'community of practice' comes into being with people who are engaged in a process of collective learning in a shared domain of human endeavour; for example, a tribe learning to survive, or a group of pupils defining their identity in a school environment. They share a concern or passion for something they do, and learn how to do it better. Wenger (2012) further differentiates between a community and a 'community of practice'. He argues that a group of people living together is a community, but they are not per se a 'community of practice.' A 'community of practice' is comprised of three elements, namely the domain, the community and the practice (Wenger, 2002). Insofar as the domain is concerned, a 'community of practice' is characterised by the commitment of its members to a specific goal. The members of a 'community of practice' are identified by specific competencies, namely as those who contribute towards the attain-ment of a specific goal. Members of a 'community of practice' value their collective competencies and learn from each other, despite the fact that none of the members may value or even recognise his or her own expertise.
The second element that comprises a 'community of practice' is mutual engagement. Members of a 'community of practice' assist each other and are engaged in joint activities and discussions; they help each other and share information; they build relationships that enable them to learn from each other. Applied to this study, postgraduate students who work together and support each other in completing their research may form a 'community of practice.' Furthermore, supervisors who work together and share information with regard to the supervision process also form a 'community of practice.' As members of a 'community of practice', they are able to discuss issues of common concern, bounce ideas off against each other, and find solutions to problems experienced in their postgraduate study and supervisory journey. This 'community of practice' is also of benefit to novice supervisors who usually work in isolation with no assistance from experienced supervisors. Since the Community of Practice Theory exists on the relationship of equals, members of a 'community of practice' are all able to learn from each other (Wenger, 2012).
The third element characterising a 'community of practice', is the existence of shared repertoires. Members develop shared repertoires of resources and ways of addressing recurring problems. This relationship of mutual interest and dedication to a common goal develops over time (Wenger, 2012).
The three elements discussed above form a 'community of practice', and it is necessary that these three elements need development at the same time to create such a 'community of practice.' In the section below the researcher discusses the implications of the Community of Practice Theory with regard to student supervision at an ODeL institution.

Student Supervision at UNISA and the Community of Practice Theory
In this study, the researcher argues that the majority of postgraduate students at UNISA, a major ODeL institution, fail to complete their studies or drop out of their studies due to both academic and nonacademic reasons. Non-academic factors include health, work-related and family problems, to mention but a few (Bitzer & Albertyn, 2011). Academic factors include the following challenges: namely most of the postgraduate students at UNISA are English second language speakers, and many of them also fail to fully understand what is required of them (UNISA, 2008). Furthermore, many live in geographical and intellectual isolation, as they often reside in rural areas with no libraries in their vicinity, and with internet connectivity either lacking or unreliable (Craig, 2015). These factors inevitably lead to the students' frustration, high dropout rates, and their inability to complete their studies within a reasonable period of time (Letseka & Pitsoe, 2014). Within this context, the formation of a 'community of practice' would provide the students with opportunities to form strong networks, which would serve as support structures to achieve a common goal. These networks could also have the potential of continuing long after the students have graduated. However, many conditions, also prevalent at other ODeL institutions, hinder the creation of the necessary 'communities of practice' between both the students (especially those residing in rural areas) and their supervisors. In this challenging context, the quality of supervision, a change in institutional culture, and supervisor competence in using different technologies to assist the students are essential.
Below the researcher will discuss the research methods used in conducting this study.

Research Method
By means of purposive sampling, the researcher selected ten postgraduate supervisors from the College of Education at UNISA, an ODeL institution in South Africa. The participants were responsible for postgraduate supervision in different departments within the College of Education, and thus represented different sub-disciplines within education. Of the ten participants selected, five were males and five were females. Six participants had more than ten years' teaching experience at a teacher education institution; the remaining four academics had five years' teaching experience at university level, and at least ten years' teaching experience at either primary or secondary schools. All possessed doctoral degrees, completed at different dates. The participants also varied in their experience of supervision, ranging from 20 years and more to less than five years in their rankings, which ranged from the position of professor (2), associate professor (4) and senior lecturer (4). The researcher recruited all these participants personally and met with them in their offices, where the researcher explained the purpose and procedures of the project to each one of them. The researcher emailed them a letter of consent after they had agreed to participate in the study, together with a questionnaire with an indication of the return date. Although the researcher initially contacted 15 academics, only ten returned the questionnaire.
An open-ended questionnaire was the means of collecting the data, consisting of nine questions to which the participants had to respond freely in their own words. The first section of the questionnaire required the participants to give their biographical information, namely their quali-fications, years' of supervisory experience and the number of master's and doctoral students supervised to successful completion. The open-ended questions included, among others, the manner of allocation of the students, the procedures used to guide the students through the research project, and the type of feedback provided to the students. No limit was set on the length of the responses and there were no predetermined options. Thus, the questionnaire acted as a writing prompt for the participants. The intention of the questions was to probe the supervision experiences and concomitant challenges at an ODeL institution. The researcher piloted the questionnaire beforehand with two academic supervisors at another university, who did not form part of the study, in order to test its efficacy. Ethical considerations included, among others, stating the research aim, indicating voluntary participation, ensuring anonymity with pseudonyms, together with the right to withdraw from the study at any time without penalty. In order to allow for member checking, the participants were given the findings of the study in order to verify if their responses were captured accurately.
The researcher made use of the following stages of data analysis, namely the initial stage that involved the contextual coding of the data. The researcher reviewed the data from the questionnaires to identify the frames of analysis, which are levels of specificity within which the examination of the data took place. These frames of analysis demarcated segments within the data. Each segment received a label with a 'code'a word or phrase suggesting how the segment informed the research question or research objective. The coding of the data resulted in the formation of categories. Guided by the main research question, the researcher analysed each questionnaire item for concise thoughts and categories. Through an interpretive process, patterns and trends emerged from the data, grouped into broad themes. At this stage the researcher prepared the overall narrative through summarising the prevalence of the patterns, and trends, discussing similarities and differences between them and comparing the relationship between one and more of these. Although the initial research question and the theoretical framework suggested some of the expected categories, the researcher was open to categories and themes that emerged from the data. The four criteria for trustworthiness for the insurance of trustworthiness were namely, credibility, dependability, confirmability and transferability (Shenton, 2004).
The credibility of the research results was obtained through the researcher familiarising herself with the culture of the research participants before the commencement of the research project. The researcher also engaged with the participants in order to gain an in-depth understanding of their work environments and to establish a relationship of trust between them and the researcher. The researcher further ensured the credibility of the results by informing the participants of their rights not to participate, in order to make sure that those participating in the study were willing to offer the information needed. Member checking was done to check the accuracy of the data at the end of the data-collection sessions. Dependability was based on the provision of detailed information of what the data-gathering instruments entailed, as well as evaluating the effectiveness of the process of inquiry.
The researcher also provided a detailed description of the method used for data collection and analysis to allow for the determination of how far the data and the construct emerging from it may be acceptable, which then confirms the conformability of the research. Lastly, the transferability of the results depended on the provision of sufficient contextual information about the fieldwork sites.
In the section below a discussion of the findings of the research will follow.

Discussion of Research Findings
Nine major themes emerged from the analysed research data, which will be discussed in the section below.
They are the following: a. procedures for the selection of the postgraduate students; b. the supervisors' first responses to the postgraduate students; c. the supervisors' experience of supervision; d. successful supervision processes of the experienced supervisors; e. advancing realistic supervision steps; f. guiding students through the structure of the thesis; g. encouraging the postgraduate students to be diligent; h. difficulties encountered in an ODeL model with regard to struggling students; and i. the supervisors' many roles.

Procedures for the Selection of the Postgraduate Students
The results of the study indicated that the participants were not responsible for recruiting their own students. The academic files of the students already admitted to the university's admissions office system were handed to the participants without the supervisor's knowledge of the students' academic background and unique needs. The supervisors were thus obligated to accept the students, irrespective of the students' research focus. The evidence is the following quotations from one of the participants: "The research co-ordinator allocates students according to the research niche of the lecturer and the student's topic." "When the chairperson of the department's Higher Degrees Committee signs off the M & D [i.e., master's and doctoral] applicants, he immediately allocates the supervisors as a condition for the students' admission." "The departmental research co-ordinator allocates students to the lecturers according to their area of specialisation." "As a small department we are forced to take students in areas that we do not have any knowledge of in order to receive funds for the department. So, the research coordinator just gives you what is available and you learn with the student, which is hard." The findings indicated that the supervisors at this distance e-learning institution were not involved in the selection of the postgraduate students. As a result, they did not know what their students were capable of, and what their needs were. This made it difficult for both the students and the supervisors to establish a sound relationship, which is important in the supervision process (Manathunga, 2012).

The Supervisors' First Responses to the Post-Graduate Students
The second theme that emerged was the initial response of the supervisors to their postgraduate students. The following quotations are relevant: "I outline my expectation on my initial communication and promise that I will give my support throughout. I provide my email address and telephone number and I immediately update my information on myUnisa regarding student activities, if there are any." "I first send an e-mail to introduce myself, and propose the timeframes within which the research should unfold. This is then followed by a dialogue [e-mail correspondence] during which the terms of reference are agreed upon." The first communication that the supervisors had with the students was to inform them by means of emails or telephonic conversations of their appointment as their supervisors. The supervisors had, therefore, to find suitable ways of introducing themselves to the students, of establishing a relationship of trust, and of determining the students' needs, in order to assist them with their studies, without having the advantage of face-to-face interaction typical at traditional institutions. Many of the students came from remote areas of the country or beyond its borders, since UNISA accommodates students from all over the world (UNISA, 2008). The importance of establishing supervisor-student rapport is in line with the observations by Manathunga (2012) and Wisker et al. (2007), who stressed the need for establishing a sound relationship with students for successful supervision. The participants reported that thereafter the students' research proposal was a means of communication between the students and their supervisors. It was at this stage that the supervisors provided the students with written feedback on their proposals. The students' subsequent revision of their proposals, based on the supervisors' feedback through in-text comments and explanatory notes, indicated to the supervisors the effectiveness of this mode of communication or the lack thereof. The researcher therefore argues that the students and the supervisors could benefit from a 'community of practice' at the beginning of the study, where both the students and the supervisors meet face-to-face to introduce themselves. It is necessary to establish the students' needs as supervisors, and then chart the way forward on how to assist them. Prospective students through a discipline-based workshop would also be able to form a group according to their field of study, exchange contact information, and raise their concerns.

The Supervisors' Experiences of Supervision
According to the findings of this study, the more experienced supervisors appeared to be effective in setting the tone of engagement, and making students aware of their role as postgraduate students quite early in the relationship. They therefore did not encounter many problems with the supervision process. The following comments from experienced supervisors serve as an illustration: "I provide my students with the programme to follow during the duration of the study." "I initiate a conversation with the student by sending him/her an email to introduce myself as the supervisor and advising him/her to email me a copy of their proposal within a certain period of time. I also use this opportunity to set the ground rules and enter into an agreement with him or her on the way forward." "The university informs the student through an email as to who his or her supervisor is. I then wait for the student to contact me, and that gives me an idea of the kind of student he or her is. I only contact him or her through an email when I realise that he or she is not going to initiate the discussion." Both the experienced and novice supervisors expressed the need for effective communication in supervising postgraduate students through ODeL. However, the procedures used for effective communication differed with regard to the content and the quality of the feedback. A 'community of practice' consisting of experienced and novice supervisors can resolve the differences in respect of communication and feedback between these two groups (Wenger, 1999). Collaboration between experienced and novice supervisors could enhance the quality of feedback and communication. This would enable the two groups of supervisors to develop shared varied experiences on what quality feedback entails, and allow them to reflect on their comments to students. Such discussions will allow both the experienced and novice supervisors to find effective ways of addressing recurring problems (Wenger, 2012).

Successful Supervision Processes of the Experienced Supervisors
The experienced supervisors were able to determine whether the students understood what was Manyike required of them (Lessing, 2011). Most experienced supervisors reported that they required from their students to provide them with written feedback after each consultation. Face-to-face consultations were, however, only possible with students who live close to the university, or with those students who were willing and able to travel to the university. In this regard, communication was not strictly ODeL only. The supervisors also expected that the students stipulate times when they expected to have completed their work. The following quotations capture this succinctly: "I give them my schedule to enable them to know when I may expect feedback and when I will be on leave." "I always send them reminders of when to submit feedback at least three weeks in advance." "I make sure that the first chapters are perfect and that the referencing is also correct. Accordingly they know from the start what my expectations are." "I always return the work that has not been edited and instruct them to have their work edited before they submit it to me." These arrangements allowed the experienced supervisors to plan their schedules and to advise the students if the suggested periods were realistic and if not, to suggest alternatives. Since experienced supervisors are successful in respect of supervising the aspirant students the novice supervisors needed to work closely with them, forming a 'community of practice' (Wenger, 2012). The collaboration between experienced supervisors and novice supervisors would enable them to share ideas amongst themselves and further enhance the novice supervisors' knowledge in dealing with different aspects of supervision.

Advancing Realistic Steps in the Supervision Process
Although postgraduate students are expected to actively participate in their own learning, it could be enhanced if they are given the opportunity to work in groups, and thus establishing a 'community of practice.' Most of the ODeL postgraduate students are adults and mid-career professionals who study in order to improve their professional opportunities (Wisker et al., 2003). Hence they need to set reasonable time frames for their studies that accommodate the demands of their professional lives.
In an ODeL institution, the written feedback from students on the meetings held with the supervisor also assists the supervisors to identify the students' needs with regard to their proficiency skills in terms of their academic writing. The meetings held with supervisors are important, because most of the postgraduate students registered at UNISA are English second language speakers with varying levels of language proficiency, and they come from diverse cultural backgrounds (Letseka & Pitsoe, 2014). The novice supervisors in the study indicated that they gave the students detailed feedback. They, however, expected the students to take the initiative of contacting them when there was something they did not understand. The novice supervisors indicated that they waited for the students to submit their work and only reminded students to submit their work if they did not hear from them for a long time.
The novice supervisors' responses make it apparent that they need to work with the experienced supervisors to learn from them what effective communication in an ODeL context entails. There exists a need for the establishment of a 'community of practise' (Wenger, 2012) among both the novice and the experienced supervisors in order to strengthen the supervision process. The success of postgraduate students depends mainly on the personal relationship between the supervisors and the students (Mouton, 2011;Wisker et al., 2007). This relationship can only be realised through effective communication between both the students and the supervisors, and that they all are able to meet the agreed-upon times.

Guiding Students Through the Structure of the Thesis
The supervisors who participated in this study all agreed that it was important to give detailed guidance to ODeL postgraduate students, who mostly appeared not to know what the supervision process entailed. The lack of knowledge of the supervision process illustrates that most postgraduate students need workshops for guidance. All the supervisors should participate in such workshops, as this would also assist the novice supervisors to identify the students' needs and to develop their own supervisory skills. Postgraduate students could also collaborate with their peers, and this would ease their feeling of isolation and improve the quality of their work. All the participants considered it important to provide the students with the structure of a thesis.
The following quotations give an indication of the various strategies used by supervisors to provide postgraduate students with the structure of the thesis: "I developed three documents: the format of the research proposal; the structure of the research proposal; and a mock tutorial letter on how to develop a critical literature review. These documents become the basis for the student to comply with the development of a coherent research proposal." "I scaffold the work and give them the length of the proposal sections especially with regards to the development of the proposal so that they do not become confused." "I email a sample proposal to the students and advise them to read through it and follow the same structure because it works well." "I tell them to go to the library and read at least five dissertations and theses on their topic to get an

idea of the required format. This is very important because most of the students are not familiar with how to conduct research. This sort of thing gives them an idea of what is expected of them."
From the above it is clear that the supervisors were aware of the pitfalls the students may experience. Although UNISA provides the students with written information in respect of the format of the proposal, the reference style, as well as other relevant information to assist them in their studies, these documents are either not read or the students fail to grasp and apply the contents. The intention of ODeL is to provide access to education, especially for previously disadvantaged persons who may not be able to pay the exorbitant fees asked by traditional face to face institutions (Letseka & Pitsoe, 2014;UNISA, 2008). If the intention of opening access to education is to be realised, it is essential that ODeL institutions ensure that the students have access to the internet to access information. Internet access will enable postgraduate students to participate in group discussions, and in so doing they will be able to form a 'community of practice' (Wenger, 1999) and improve the quality of their work. Novice supervisors, who are mostly young, can also assist experienced supervisor as many novice supervisors are more comfortable with using internet-based technology. Many UNISA students are academically under-prepared for postgraduate studies, and supervisors need to make up for this unpreparedness to graduate studies to ensure successful completion through remedial work at the beginning of the year in workshops. The research participants indicated that they assisted their students by using the following strategies: scaffolding, providing a good sample proposal to emulate, and referring students to the librarians to assist with literature retrieval. Supervisors working individually with students mainly use these strategies. However, this could be done through collaboratively creating a centralised learning community, instead of working with postgraduate students individually.
Working collaboratively is in line with Manathunga's (2012) observation that for supervision to be successful, explicit instructions should be given to postgraduate students. Many students at UNISA complete their undergraduate studies through distance learning and have no experience of face-to-face traditional institutions. Therefore, they require guidelines with regard to the use of the internet and literature search for their studies and planning due dates for the regular submission of their work. Experienced supervisors appeared to be able to provide students with the structure of the thesis as well as with realistic schedules which further assisted students in planning their work and calls for a 'community of practice' among supervisors to work collaboratively and share their experiences.

Encouraging the Post-Graduate Students to be Diligent
According to Malan et al. (2012), postgraduate students have to understand their role as students and that of their supervisor in order to succeed in their studies. Below are some of the supervisors' comments: "I give deadlines in emails and myUnisa when I send their work back to them for corrections. I remind them about submission dates a week or two weeks before." "It starts with the research proposal where I encourage them to treat it as an assignment and to complete it within three to four weeks. On the approval of the proposal, we develop a working plan that requires a draft chapter to be developed and submitted every month. A fortnight into the writing I send them a friendly reminder, as per agreement namely, "Two weeks gone; two weeks to go." "I remind them to submit their work when I realise that I have not heard from them after a month." "I call them and tell them to submit their work after three weeks of not hearing from them." The above quotations show that experienced supervisors are able to plan the project with the students and set agreed upon time frames. The agreed upon timeframes enable the students to plan their work and work diligently. An example can be gleaned from an experienced supervisor who wants chapter submission on a monthly basis and does not wait for the end of the month to remind the students. The initial reminder also serves as a prompt to students to communicate the problems they might be experiencing. What can be gleaned from the quotations above is that these academics are not conducting group supervision, but individual supervision. This is not in line with the 'community of practice theory'. I therefore argue that experienced and novice supervisors could work collaboratively in structuring the studies and as the success of the study project depends mainly on the ability to plan the project well, a skill which most novice supervisors lack.

Difficulties Encountered in the ODeL Model with regard to Struggling Students
As mentioned before, UNISA has recently experienced pressure to admit more students, including previously disadvantaged students, who are mostly academically disadvantaged and who are English Second Language (ESL) speakers (Koen, 2007;Lessing, 2011;Mouton, 2011). In South Africa, a country that shares the features of both a developed and a developing economy, most UNISA students are located in isolated rural areas, without basic essentials such as internet connectivity and computers (DHET, Republic of South Africa, 2012). Internet connectivity is something the University and the government have to address if they want to empower people around the country through education, namely to strengthen the use of technology which is essential for successful completion of postgraduate studies as well as to prepare the youth for the world of work. Ensuring the availability of internet connectivity and computers by the university and that they work properly is fundamental to the offering of ODeL. Furthermore, most of the postgraduate students from the rural areas lack the basic academic English language skills to cope with the demands of postgraduate education (Koen, 2007). The students' proficiency in English as an academic language may improve by means of ODeL. Researchers such as Lemmer and Manyike (2012), have confirmed the link between language proficiency and academic success within the South African context. UNISA as an ODeL institution should address students' language proficiency skills, as it determines their success in postgraduate studies.
In respect of the above situation, the participants indicated that they were acutely aware of the drawbacks of supervising students who were academically under-prepared due to inadequate proficiency in English, access to technology, and technological skills. Often when these students register at UNISA for postgraduate studies, they not only struggle with the content they have to master, but they also struggle with learning the new technology, which is an essential tool for their academic success. All the supervisors of postgraduate students who participated in the study were concerned about the fact that most students were unable to express themselves in English, as the official language of instruction. In their everyday communication, most of the students used their indigenous languages and used English primarily for official purposes. Although most of the postgraduate students in education are teachers and use English as the language of teaching and learning in their classrooms, the reality is that they frequently tended to code-switch (Planas & Setati-Phakeng, 2014). Since postgraduate study requires extensive reading and writing, the inability to read advanced academic texts with enough understanding and to write coherently leads to academic failure (Mouton, 2011). Most of the supervisors indicated that they struggled to understand the students' work, and that the students did not fully understand their feedback. Academic writing is a challenge even to students who are English first language speakers, but these challenges are more profound in respect of students from diverse cultural backgrounds and impoverished communities (UNISA, 2008). The following quotations capture the views of many of the supervisors with regard to the quality of the students' work: "Some students … need a verbal or oral explanation and a face-to-face feedback session." "Some students fail to understand the feedback provided and I make an appointment with them to see them during the school holidays if they stay far from the university campus." "When they are 'lost', they call me and I am able to explain again." "When the student submits the work and I struggle to understand what was written I return the work to him/her." "At times the resubmitted work is still as bad as the first submission. I just compare it with the first submission and when I feel the corrections done are not enough, I return the work to the student with explicit instructions that he or she should give their work to someone to edit." Experienced supervisors indicated that they often return the student's work after reading only a few pages when they noted a lack of coherence in the argument, and an unacceptable number of grammatical errors. They usually give the students clear instructions to have their work edited by a language editor before resubmitting it. Experienced supervisors were also able to correct the content and the technical aspects. They claimed this was important from the very outset to 'train' the new students in the rudiments of referencing style and the bibliography. They reported that, on occasion, they had to return a student's first chapter more than three times in order to teach him/her the correct use of the technical skills. This is in contrast to the practice of novice supervisors who tended to concentrate more on content and to ignore the technical details, which created further problems as the study neared completion and in view of submission for the examination.
The supervisors agreed that not all the students understood what was required of them by means of the in-text feedback provided. They also agreed that they sometimes had to call the students to the campus for face-to-face interaction. Face-toface interaction was easier for students who were South African residents as opposed to those living abroad. Given the fact that UNISA is an ODeL institution, face-to-face interaction between the students and the supervisors is not always possible. The most common mode of personal interaction is through email, Short Message Service (SMS) or telephone. However, since the introduction of video-calling applications such as Skype, which allows a degree of virtual face-to-face interaction, the supervision of postgraduates is now much easier. This very important solution can change the support to students drasticallygiven that they have the necessary connection for such communication. However, video calling requires highspeed internet connectivity. In developing countries, internet connectivity is still prohibitively expensive, and it is only readily available within urban areas. Therefore, the rural students have to travel to urban centres in order to access high-speed internet connectivity. Nevertheless, the supervisors should be encouraged to optimise video calling applications, although only a few participants admitted to using them. Their use can enhance the students' learning and assist both parties in knowing and understanding each other better, given the importance of establishing a 'community of practice' to ensure successful postgraduate supervision.

The Supervisors' Many Roles
All the supervisors who participated in this study indicated that their role was not only to assist their postgraduate students to succeed academically, but that they were responsible for their students' overall wellbeing. They indicated that the supervision of both master's and doctoral studies involves a long-term commitment and intensive work, and in many cases, the formation of a lifetime professional relationship with the students. It is, therefore, the responsibility of the supervisor to fulfil the many roles of confidante, friend, mentor, counsellor, career guide, and of financial advisor, namely to advise the student about available funding in the form of student grants.
Below, the researcher will elaborate on the supervisor's role as role model and financial advisor in the light of the data received.
The supervisors who participated in this study were aware of the different roles that they had to play in assisting their postgraduate students to succeed. Four of the experienced supervisors invited their students to attend conferences and to present papers at these conferences in order to socialise the students into the research community, which resembles Wenger's (1999) 'community of practice.' This procedure is in line with Mouton's (2011) argument that experienced supervisors possess the ability to provide guidance and structure to their students. Mouton (2011) furthermore observed that supervisors have the ability to detect, at an early stage, whether or not a student will be able to complete the study. Most experienced supervisors were also able to provide their students, especially those in rural areas, with financial assistance by using their own grants, in addition to assisting the students to obtain study grants. Experienced supervisors were also able to model good writing by providing their students with their own articles as examples of good writing skills (Albertyn, Kapp & Bitzer, 2008;Bitzer, 2011). The novice supervisors also fulfilled their role as confidante, where they indicated that the students tended to confide in them as an excuse for not submitting their work on time. They often presented the excuse of not being able to submit their work on time due to ill health or family issues.
The following quotations provide evidence of the above.
"I am a writer and researcher who travel widely disseminating my ideas at international conferences, and my work often gets published in international journals belonging to the host associates. I share this information with my students and encourage them to become members of such communities of scholars." "Yes, I encourage my doctoral students to apply for posts at universities, especially those students who are interested in an academic career." "One of my doctoral students is now my postdoctoral student and we have a close professional relationship." The experienced supervisors indicated that they kept in touch with most of their students long after the completion of their studies. The personal relationship that they developed with the students was a rich source of job satisfaction. Both the experienced and the less experienced supervisors indicated that they were impressed with their students' personal growth and were able to learn from them, especially the more talented ones. The ability to nurture a collegial relationship, which continues after the completion of the degree, is in line with Wenger's (2002) theory of 'community of practice', where people with the same interests continue working together in order to achieve a common purpose.

Conclusion and Recommendations
The results of this study indicated that experienced supervisors were very successful in reducing the distance in an ODeL institution through the effective management of the process. They were able to detect problematic areas early and provided their students with the necessary support. Most of the experienced supervisors provided their postgraduate students with the structure at the beginning of the study to allow for effective planning and further engagement. However, the novice supervisors were not as able to detect problems and, as such, stumbled along through the process. They were unable to provide their postgraduate students with the required structure, time management skills, technical help and financial support where necessary.
The results of this study further revealed that the experiences of ODeL postgraduate supervisors were varied, and depended on both the number of years of supervisory experience as well as the type of mentorship they had received as academics. The researcher recommends that experienced supervisors should conduct workshops to assist novice supervisors with regard to effective supervision procedures. Forming a 'community of practice' between experienced supervisors and novice supervisors can support the process of supervising, ideally a dual overlapping community, a community of supervisors and a community of graduate students. Mentorship programmes can be introduced to support novice supervisors in order to improve their supervision capacities, thus contributing to the economic development of the country through the production of well-qualified human resources needed for international competition. It is recommended that workshops be conducted for both supervisors and students, and that a strong relationship be established between novice and experienced supervisors in order to share their supervisory experiences. Funding for biannual meetings and seminars be provided by UNISA, where students and supervisors can congregate for a specific session, not a one-size-fits-all workshops, as is commonly the case. Furthermore, there is a need for uniformity of supervision procedures within the same institution. ODeL universities should explore various ways of subsidising postgraduate students, especially those residing in rural areas, with cheaper high-speed internet access can be explored by ODeL institutions. Finally, UNISA need to establish an office to which postgraduate students can bring supervision challenges as well as a 'community of students' (online), where they can share their challenges and successes with university officials to obtain advice and share resources. A closed Facebook group is an example of this kind of virtual community, where students can express their dissatisfaction with the supervision process as this may allow for the early detection and resolution of problems, and reduce the dropout rates.

i.
Published under a Creative Commons Attribution Licence.